Productising an AI model for instant hiring matches

The challenge:

The founder had developed a strong AI matching model, with early results demonstrating its potential to identify candidates who progress to interview. The opportunity was to build the product around that intelligence - translating a compelling vision and sophisticated model into an intuitive, trustworthy experience for both founders and candidates.


Our approach:

We worked across discovery, product definition and experience design - translating customer evidence into product logic, workflows and development priorities. Our approach included:

  • Researching founder and candidate behaviours, needs and decision-making challenges

  • Defining the product proposition, core principles and MVP priorities

  • Translating research into end-to-end B2B and B2C workflows

  • Designing interactive prototypes to define and demonstrate the core product experience

  • Producing product requirements and a prioritised roadmap to support development


Outcomes:

  • Transformed a high-performing AI model into a clearly defined and testable product proposition

  • Established coherent journeys connecting founder needs, candidate information and AI-supported evaluation

  • Identified the evidence, transparency and human oversight needed to build trust in recommendations

  • Created a central product experience repository bringing together the information architecture, user flows, interactive prototypes and key experience decisions

  • Created a prioritised roadmap connecting customer needs, product opportunities and development

  • Created a practical foundation for the team to develop, test and bring the product to market


Impact:

The work connected the founder’s vision and AI model with a coherent customer experience - helping founders make better-supported hiring decisions while enabling candidates to understand how their experience is assessed through a recruiter’s lens.


The UK recruitment industry contributes £40.6 billion to the economy, (REC).